REW for the first time

[LCRLive]

So in other words, applying room calibration the same way we applied soundcard calibration so that it factors in and takes away what the room is doing to measurements. If a soundcard calibration can exist why can't a room calibration file to account for everything?

Basically a room correction calibration file is badly needed within this and other software like the Dayton one because most of us cannot afford or allocate space to build very large anechoic chamber testing rooms.

On soundcard calibration, this pic is what my pro 40 is doing. I was expecting this rack unit to be a lot cleaner than this.

Also wondering if there is some latency in the system, wont that completely screw up the measurement with some sort of phasing? or is the signal path for this thing properly compensating for all that?

 

[John Mulcahy]

The room's effect are in the time domain as much as the frequency domain, there is no static frequency compensation that can remove them. Reflections from the room's surfaces arrive later than the direct sound, so the primary method is to window the impulse response to exclude those reflections. The trade-off is that the shorter the window, the poorer the frequency resolution, so measurements need to be made with speaker and mic as far from room surfaces as possible. Measurements are sometimes made outside on elevated platforms for that reason.

There is always latency, even if it is only the time for the signal to travel from speaker to mic. It can be estimated and removed ("Estimate IR Delay" in the graph controls). A timing reference (such as a loopback connection on a second channel) allows consistent relative timing between measurements, the Measure dialog has a field to enter a timing offset figure to apply to measurements.

 

[LCRLive]

The room's effect are in the time domain as much as the frequency domain, there is no static frequency compensation that can remove them. Reflections from the room's surfaces arrive later than the direct sound, so the primary method is to window the impulse response to exclude those reflections. The trade-off is that the shorter the window, the poorer the frequency resolution, so measurements need to be made with speaker and mic as far from room surfaces as possible. Measurements are sometimes made outside on elevated platforms for that reason.

What if I put large jenga-like random piles of foam on the floor in between the Mic and the speaker and replaced the drop ceiling tiles with special acoustic ones? (With a bunch of foam on top in behind up there too).

 

[sam_adams]

What if I put large jenga-like random piles of foam on the floor in between the Mic and the speaker and replaced the drop ceiling tiles with special acoustic ones? (With a bunch of foam on top in behind up there too).

To achieve results like this:

you will need several feet/meters of absorbing material on every surface in the room. This is not practical in rooms of average size. The above is the measurement of a single driver that was done in a very small room that has eighteen inches of dense polyester fiber fill on all the walls (floor to ceiling) and ceiling of the room. The measurement is 1M with no smoothing applied. The dips in response at 700 Hz and 1200 Hz were determined to be cancellations from the front and back of the baffle the driver was mounted on. The driver has pronounced breakup starting at 4500 Hz.

 

[ddude003]

Along the lines of what @sam_adams suggests above... Look at the specs for a Vocal Booth... Google is your frenemy...
Basically a room 6x6x6 or 6x8x6 with acoustic absorption completely covering all and every surface... Not some random foam placement... Basically flat and dry from 20Hz to 20kHz and a reverb (RT60) of 0.2 seconds...

 

[DanDan]

Many many people have used REW for the first time. Many many of the questions are the same as it ever was....... They have all been answered a gazillion times.

Some guides to REW and acoustic measurement

AustinJerry's guide has good coverage of using USB microphones.

www.avnirvana.com

 

[LCRLive]

So once again I have been told by Parts Express that their Dayton measurement kit does room compensation for all the above discussed parameters but it only does so within a certain window so they still recommend treating as much as possible and with what i said I'm going to do in the room they said thats more than enough).

Although I dont know how the hell that is even possible because if you send a sweep with a particular speaker to get the room EQ, the how can the sweep itself be set to account for the speaker's own frequency curve? And same question goes for the actual room testing. If one is sending a sweep, we are at the limitations of the speaker's own inherent EQ curve in that sweep send so it would appear that even testing rooms would be compromised if a person is sending the sweeps with speakers which do not have perfectly flat frequency characteristics.
Which is now laughable in the most visceral sense of the term as most people have probably been running compromised tests.

 

[JStewart]

And same question goes for the actual room testing. If one is sending a sweep, we are at the limitations and test speaker's own inherent EQ curve in that sweep send so it would appear that even testing rooms would be compromised if a person is sending the sweeps with speakers which do not have perfectly flat frequency characteristics.

Uness you consider that the speaker and room (and the electronics in the signal chain) are working together as part of a system and it is the sound produced by the system you are measuring and correcting for.
Or am I not understanding your point?

 

[DanDan]

LCRlive said
Which is now laughable in the most visceral sense of the term as most people have probably been running compromised tests.

Seriously?
Abe Lincoln:- "Better to remain silent and be thought a fool than to speak and to remove all doubt"

 

[DanDan]

@LCRLive. I thought a little humorous jibe might help some cop on, but unfortunately not. You represent The Holy Trinity.... Loud Confident Wrong. Apparently you think most of us here are under some delusion, which you like to laugh at. Have you seen The Joker? In reality it is your fundamental ignorance of what is at play which is not funny, but sad. So ignoring humanity, to address your lack of technical understanding..........+1 to SStewarts effort to point you towards reality. I will add that the location of the speaker in the room changes how the room is driven, and indeed the room does also modify the output from the speaker to a smaller extent.
Most if not all measurement involves reference to a close tolerance standard. e.g. Metres of stone or platinum, or the original Kilo. in Paris, Atomic clocks broadcasting on the airwaves. Here, a Hi Fi forum, most of the speakers will have close tolerance and full range. Say +/- 3dB 30Hz- 20K. . Typical room anomalies can span 30dB and more.

 

[LCRLive]

And

@LCRLive..... I thought a little humorous jibe might bring you to your senses..... Unfortunately not. Apparently you find the delusion which most of us here are under to be laughable. In reality it is your fundamental ignorance of what is at play which is not funny, but sad. So ignoring humanity, to address you lack of technical understanding......
Most if not all measurement involves reference to a close tolerance standard. e.g. Metres of stone or platinum, or the original Kilo. in Paris, Atomic clocks broadcasting on the airwaves. Here, a Hi Fi forum, most of the speakers will have close tolerance and full range. Say +/- 3dB 30Hz- 20K. . Typical room anomalies can span 30dB and more.

By you indicating a 30 Db or more variation, (why did you not write +/- 15 when you wrote +/- 3db for speakers?) you are including very acoustically unfriendly-shaped untreated rooms as a very invalid reference for this compare.

After a room is treated with panels and bass traps, a decent room would have a 15 db or less Window from which to begin to fine tune more professionally.

If one of the moderate to higher end builds, the +/- 3db window on speakers is about right. But I've seen even larger windows on stuff priced insanely high.

Compare to a +/- 1-ish db on an Adam S3V official measurements and one starts to realize that lots of snake oil seems to be greasing up the price tags in the speaker industry.

 

[DanDan]

why did you not write +/- 15 when you wrote +/- 3db for speakers?

Because- when two coherent signals sum, the peak is +6dB. When they cancel the null is zero. So, in a room cancellation dips are way bigger than peaks. Also the extents of whatever dB window is very much dependent on what smoothing is applied. I don't really get whatever point you are trying to make with the ADAM but I will point out that such extracts are pretty meaningless without considering the off axis responses. +/- 1 dB on axis strikes me as fanciful, I can't see ADAM claiming such nonsense. If you have questions they are welcome. But this strategy of enquiry by accusing everyone of delusion or ignorance is perverse. The Inquisition didn't work. Enough, or the naughty step.

 

[arivel]

HI .
I have two questions to ask . I apologize if the answer has already been given somewhere else in this forum.
Is it possible to use the sound card to check if the amplifier works correctly over the entire frequency range or if it has problems? .
the second question refers to the following intervention

indirect method to listen to the sound of other people's speakers

Hi everyone . I would like to discuss something together. using software and microphone like REW and with sweep or pink noise it is possible to publish graphs such as frequency response, distortion, harmonics and many other data. unfortunately, however, only the most skilled and with a lot of...

www.diyaudio.com

I haven't reread all the posts but I remember that they talked about an anechoic chamber. it is difficult to afford to build an anechoic chamber, it would be something to be able to do it using a normal room. I thought I'd publish it here too because there are people who are experts in measurement and equipment dedicated to it who might have some ideas on the matter.
it is true that the effects of the two rooms add up but if the person who receives the file is on the other side of the world he could
make use of a studio monitor listened outdoors. What do you think it would be like?

 

[ddude003]

To your second question... You might google room impulse response convolution reverb... And speaker cabinet impulse response... And waves abbey road studio 3 plugin...

 

[arivel]

hello and thank you. I have another question to ask.
I've been working on a Karlson prototype-variant for some time.
I hear a box effect when I listen to music, I wonder if REW can help understand the cause of this.

 

[DanDan]

Speakers can be measured anechoically, almost, up on a pole outdoors. Headphones provide easy access to anechoic listening. REW can certainly measure the FR and Distortion etc. of an Amplifier or other electronic device. Sorry I am not familiar with Karlson.

 

[arivel]

HI .
I got tired of taking measurements by closing my ears with my hands or using earplugs and I would like to buy headphones.
better those used in the workplace to soundproof from noise or those for studios?.

 

[JStewart]

HI .
I got tired of taking measurements by closing my ears with my hands or using earplugs and I would like to buy headphones.
better those used in the workplace to soundproof from noise or those for studios?.

If the REW sweeps are loud enough to bother you, perhaps they’re too loud. Having the sweep volume at your normal safe listening level should work. Usually I’ll measure just above the level of normal conversation, which is >20dB above the noise floor in my case.

Another alternative for you might be to set a sufficiently long start delay on the REW measurement window so you have time to leave the room.

 

[arivel]

every time I take a measurement I click on the check levels button to check that everything is ok, if the volume is set too low the low level message appears. From this message I assume that REW is unable to make a reliable measurement in these conditions. furthermore in the guide it is written to measure with a volume of 75db

 

[JStewart]

furthermore in the guide it is written to measure with a volume of 75db

That is what I said. I said

Usually I’ll measure just above the level of normal conversation,

Human conversation is 70dB on average, “just above” ~75dB.

What I don’t understand is why hearing protection would be needed or wanted at that level and so I thought perhaps you were running REW sweeps at a higher volume than you realize. It’s entirely possible for REW to NOT read the absolute SPL level correctly for a variety of reasons. There are numerous sound pressure level apps for smartphones that are close enough to accurate to double check REWs SPL reading.

If I misunderstand your situation and why you use hearing protection with REW, I apologize for interrupting.

 

[arivel]

why in the distortion graph do the THD value and the harmonic values vary depending on the zoom set?.
what should I do to display the exact values?.

 

[John Mulcahy]

The figures shown are for the position of the graph cursor.

 

[arivel]

HI . I listened to music with my K12 but some frequencies are difficult to hear.
I would like to apply a high pass filter to eliminate all the frequencies below the impedance peak Fs of the 12" woofer. If it is possible I would like to do it digitally with Windows. If this is not possible then I ask you if I can apply a capacitor and resistor between the sound card output and the power amplifier input.

 

[ddude003]

You could probably use something like the Hang Loose Host https://accuratesound.ca/hang-loose-host-hlhost/ along with a decent Equalization plugin that would contain a nice high pass filter... Or, create a FIR filter and use a convolution engine like the Hang Loose Convolver also found at the above website...

 

[John Mulcahy]

Equalizer APO is another option.